Models
In this demo, we present rhythm-only, pitch-only, and timbre-only conversion results using the following models
- SPSP-S: SpeechSplit model with small bottleneck
- SPSP-L: SpeechSplit model with large bottleneck
- SPSP2-S: SpeechSplit2 model with small bottleneck
- SPSP2-L: SpeechSplit2 model with large bottleneck
Rhythm-only Conversion
| Source Speech | Target Speech | Conversion | |
|---|---|---|---|
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L |
Pitch-only Conversion
| Source Speech | Target Speech | Conversion | |
|---|---|---|---|
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L |
Timbre-only Conversion
| Source Speech | Target Speech | Conversion | |
|---|---|---|---|
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L | |||
| SPSP-S | |||
| SPSP-L | |||
| SPSP2-S | |||
| SPSP2-L |